tc39 / proposal-bigint-math

Draft specification for supporting BigInts in JavaScript’s Math methods.
https://tc39.es/proposal-bigint-math/
BSD 3-Clause "New" or "Revised" License
37 stars 2 forks source link
bigint math proposal tc39

BigInt Math for JavaScript

ECMAScript Stage-1 Proposal. J. S. Choi, 2021.

Description

(A formal draft specification is available.)

BigInts are important for a myriad of mathematical, financial, scientific, and timing applications (such as in the Node.js process.hrtime.bigint API), and they have been therefore a valuable addition to JavaScript since their standardization in ES 2021.

Several built-in Math functions would make sense with BigInts, yet they still do not support them; they only support regular floating-point JavaScript Numbers. This proposal extends those functions’ behavior to accept BigInts:

abs\ sign\ sqrt\ pow *

min †\ max †

* pow does not accept mixed types. pow(4, 2n) will throw a TypeError.

 min and max accept mixed numeric types:\ min(0, 1n, -1) evaluates to 0,\ and max(0, 1n, -1) evaluates to 1n.\ This is well defined because < is well defined over mixed numeric types; there is no loss of precision.

When Math.min and Math.max receive values of different numeric types that nevertheless have equivalent mathematical values, then min prefers the leftmost value and max prefers the rightmost value. For example, Math.min(0, 0n) is 0 and Math.max(0, 0n) is 0n. (See issue #3.)

Philosophy

The philosophy is to be consistent with the precedents already set by the language. These precedents include the following five rules:

  1. BigInts and Numbers are not semantically interchangeable. It is important for the developer to reason about them differently.
  2. But, for ease of use, many (but not all) numeric operations (such as division / and exponentiation **) are type overloaded to accept both Numbers and BigInts.
  3. These type-overloaded numeric operations cannot mix Numbers and BigInts, with the exception of comparison operations.
  4. Some numeric operations are not overloaded (such as unary +). The programmer has to remember which operations are overloaded and which ones are not.
  5. asm.js is still important, and operations on which it depends are not type overloaded.

In this precedent, only syntactic operators are currently considered as math operations. We extend this precedent such that Math methods are also considered math operations.

Vision

This initial proposal overloads only a few first Math methods. The vision is that this proposal would open up the way to new proposals that would further extend Math with type-overloaded methods. These may include:

Excluded Math

Similarly to how some numeric operators are not overloaded (such as unary +), many Math functions that would not make sense with BigInts are excluded from this proposal. These include:

Math method Exclusion reason
acos Transcendental: very difficult to calculate when large
acosh Transcendental
asin Transcendental
asinh Transcendental
atan Transcendental
atan2 Transcendental
atanh Transcendental
cbrt No known use case
ceil No known use case; Math.ceil(3n / 2n) == 1 may be surprising
clz32 No known use case
cos Transcendental
cosh Transcendental
exp Transcendental
expm1 Transcendental
floor No known use case
fround Returns floating-point numbers by definition
hypot No known use case
imul No known use case; may complicated asm.js (see issue #9)
log Transcendental
log10 Truncation may be surprising; no known use case
log2 Truncation may be surprising; deferred to future bitLength proposal
log1p Transcendental
random No conceptual integer-only analogue
round No known use case; Math.round(3n / 2n) == 1 may be surprising
sin Transcendental
sinh No known use case
tan Transcendental
tanh Transcendental
trunc No known use case